[cp-patches] FYI: JTextField fix

2006-01-12 Thread Roman Kennke
Hi there, this fixes a regression that was pointed out by Mauve. Anthony and I have hacked together several mauve tests that show that the document property filterNewlines must be set in the setDocument() method of JTextField. 2006-01-12 Roman Kennke <[EMAIL PROTECTED]> * javax/swing/J

[cp-patches] FYI: JTextField fix

2006-02-22 Thread Roman Kennke
This adds the isValidateRoot() method to JTextField, which is a nice little optimization, because otherwise each key press would cause revalidation of the whole tree. 2006-02-22 Roman Kennke <[EMAIL PROTECTED]> * javax/swing/JTextField.java (isValidateRoot): New method. /Roman

[cp-patches] FYI: JTextField fix

2006-10-31 Thread Tania Bento
Hey, This patch fixes a minor bug in javax.swing.JTextField. The problem was that in fireActionPerformed(), when creating a new event, we were passing getText() as the parameter for the command when we should have been passing the actionCommand variable. This patch now fixes the Deepamehta logi

Re: [cp-patches] FYI: JTextField fix

2006-10-31 Thread Roman Kennke
Hi Tania, > This patch fixes a minor bug in javax.swing.JTextField. The problem was > that in fireActionPerformed(), when creating a new event, we were > passing getText() as the parameter for the command when we should have > been passing the actionCommand variable. I think there was a reason f

Re: [cp-patches] FYI: JTextField fix

2006-10-31 Thread Tania Bento
Hey Roman, On Tue, 2006-10-31 at 17:10 +0100, Roman Kennke wrote: > I think there was a reason for that behaviour. Could you please test if > possibly the new action command is getText() when the original command > was null? Ideally write a Mauve test to avoid regressions. > > /Roman Yup, you ar

Re: [cp-patches] FYI: JTextField fix

2006-10-31 Thread Roman Kennke
Hi Tania, > On Tue, 2006-10-31 at 17:10 +0100, Roman Kennke wrote: > > I think there was a reason for that behaviour. Could you please test if > > possibly the new action command is getText() when the original command > > was null? Ideally write a Mauve test to avoid regressions. > > > > /Roman >